PFMEA Workbook - ETI GROUP

ETI GROUP Six Sigma / Lean Six Sigma Lean Enterprise Business Management Systems Strategic Planning Integrated Performance Measurement

ETI Group Camas, Washington

Phone: Washington (360) 681-2188 Oregon (503)484-5979 Email: [email protected]

Failure Mode And

Effects Analysis (Process)

Process FMEAThe ETI Group © 2020 1

Performing

Failure Mode and Effects Analysis

Effectively for Processes

Albert Einstein

“Logic will take you from

A to B, but imagination

will take you

everywhere.”

Presented by:

ETI Group

Camas, WA

www.etigroupusa.com

Tel. (360) 681-2188

Or (503) 484-5979

Email: [email protected]

Copying of this material is not permitted without written approval from

The ETI Group


Contents

Learning Objectives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

What is a FMEA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

The FMEA Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Where to Focus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

Process Step/Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Potential Failure Modes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 29

Potential Effects and Consequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 33

Severity Rating. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . 37

Potential Cause . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . 41

Current Process Controls —Prevention. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . 45

Occurrence Rating. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . 46

Identifying Special Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . 49

Current Process Controls —Detection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . 51

Detection Rating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 52

Risk Priority Number (RPN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 55

Reducing Risk. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 62

Summary of Process FMEA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 68

Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

• FMEA team start-up sheet

• Process Flow Diagram Example

• Cause and Effect Diagram Example

• Fault Tree Analysis

• Quality planning with FMEA

• Manufacturing PFMEA Example

• Transactional PFMEA Example

2


Learning Objectives

• Design and implement a process to perform Failure Mode and

Effects Analysis.

• Identify and select the appropriate participants for an FMEA

Team.

• Explain the different types of FMEAs and be able to choose the

appropriate type.

• Identify potential failure modes and rank them using risk priority

numbers.

• Develop a Process FMEA through hands-on activities and

workshop exercises.

3


Introductions

• Please introduce yourself.

• Where do you work?

• What are your expectations

for today?

4


History

• Reliability Analysis technique invented by the DoD in late 1940’s to determine the effect

of system and equipment failures in terms of impact on mission success and

personnel/equipment safety.

• Became popular during 1960’s in aerospace industry

• Initial use was on improving safety – prevent accidents

• Spread to many different industries:

- Automotive

- Defense

- Electronics

- Manufacturing

- Medical Devices

- Transportation

Automotive Industry

• Adopted and adapted by automotive industry for quality improvement (in addition to

safety)

• Ford, GM, Chrysler developed standard format and approach

• Integral tool used as part of the Advanced Product Quality Planning process

• Guidelines published in Potential FMEA Reference Manual provided by the AIAG

Terminology Note: “Potential FMEA” and “FMEA” are used interchangeably.

What is an FMEA?

A Potential Failure Modes and Effects Analysis is a systematic

group of activities intended to:

• Recognize and evaluate potential failures of a product or

process.

• Evaluate the effects of failures.

• Identify actions which could eliminate or reduce the chance

of potential failure.

• Document the process & record decisions.

5


Design FMEA

• Considers potential design failures

(should consider technical/physical

limits of manufacturing/assembly

processes)

• Does not rely on process controls to

overcome design weaknesses

Terminology Note:

The terms “Design FMEA” and “Product

FMEA” are often used interchangeably.

Process FMEA

• Assumes product as designed, will

meet the design intent

• Does not rely on product design

changes to overcome process

weaknesses (but does consider

product’s characteristics)

FMEA: Design vs. Process

• Principles and steps are the same

• Purpose and objectives differ

• Rating scale definitions will differ

• Design

✓ Discover potential problems with the design of the product that will result in

safety concerns, malfunctions, or shortened life

• Process

✓ Discover potential problems related to the manufacture of the product that will

affect the product, safety, or processing efficiency

6


Results of a Design FMEA can be helpful in formulating a Process FMEA, and vice

versa.

Processes can include those for Service (such as customer/technical service, field service,

call center service, etc.).

Sketch an example of different levels for a Process FMEA in your application:

FMEA Levels

System

Sub-systems

Assemblies/Processes

Components/Activities

7


ISO/TS 16949 is an International Standard used by the automotive industry.

ISO 9001 is used by a variety of industries and serves as a basis for several industry-

specific standards such as AS9100 in aerospace.

ISO 13485 is an International Standard used by medical device manufacturers.

PFMEAs should include all operational production/service delivery processes from the

component level up through assemblies, sub-systems and system as applicable to the

organization. Other areas to include are Shipping/Receiving, material

transport/conveyers, labeling, repair/rework, service/installation, etc.

It is also recommended that the operational processes that “feed” the production/service

delivery processes undergo PFMEA ― these may include sales, design, supply chain

management/purchasing, planning, etc. Support processes such as document control,

training and metrology will also benefit from PFMEA analysis.

FMEA Applications

• Product Design (DFMEA)

• Manufacturing and Assembly Planning (PFMEA)

• Advanced Product Quality Planning (ISO/TS 16949)

• Risk Management (ISO 9001, ISO 13485, etc.)

• Safety

• Reliability-centered Maintenance

• Environmental hazard avoidance

• Problem solving (corrective actions)

• Improvement tool (preventive actions)

8


FMEA is meant as a prevention tool, and thus is ideally performed during the design

planning and review stage (and here, “design” could be the design of a product, service or

process). The goal is to prioritize development of product and process features that can

eliminate or reduce the risk of potential failures.

An FMEA is not a “one-time-event” but rather a dynamic process and learning

commitment.

FMEA can be used post mortem on actual failures (NASA performed one when the Mars

Orbiter was lost).

When to Use FMEA?

Proactive tool meant to be used when there is a:

• New design/technology/process.

• Modification to an existing design/process.

• Change in the environment, location, application or usage

profile of the design/process.

• Change in a Regulation affecting the design/process.

Should be a “before-the-event” action — not an “after-the-fact” analysis

9


The choices made early in the Design process will determine the final cost to produce the

product. Once the product reaches Production, there is very little that can be done to

impact this cost.

Why Use FMEA?

Product Costs versus Time

0

20

40

60

80

100

120

Concept Design Testing Development Production

Phase of Product Life

Lifetim

e P

roduct

Cost

(%)

Source: Design for Manufacturability:

Optimizing Cost, Quality, and Time-to-Market,

Dr. David M. Anderson

Committed Cost Incurred Cost

10


Benefits of FMEA

• Improved knowledge of the product & process

• Promotes a focus on customer requirements and consequences of failure

• Time savings, improved productivity

• Cost savings

• Reduced warranty repairs and recalls (i.e., improved reliability)

• Improved quality

• Better maintainability

• Provides documented planning history for future changes, new product and/or

process designs

• Prioritizes corrective & preventive actions for product design and process

controls

Increased Customer Satisfaction

Why Use FMEA?

With FMEA

Product LaunchTime

Num

ber

of

Desig

n C

hanges

Design & Development Process

Without FMEA

11


Customer: Internal or external departments, people, and processes that

will be adversely affected by failure.

Function/Requirement: Intended purpose of a product or process.

Cause: Means by which a particular element results in a failure mode.

Failure Mode: Way in which a product or process could fail to perform its desired

function.

Effect(s): Symptom(s) the customer will see or experience.

Consequence: What occurs in the customers’ process, system, etc. as a result of the

effects.

Controls: Mechanisms that prevent the Cause from occurring or which detect

the Failure Mode or Effect.

Severity: Assessment of how serious the consequences of the Effect will be

on the Customer.

Occurrence: Assessment of the likelihood that a particular cause will happen and

result in the failure mode.

Detection: Assessment of the likelihood that current controls will detect the

cause/mechanism of failure or the failure mode.

Terms & Definitions

Prevent Detect

CauseFailure

modeEffects Consequences

Controls

Occ

urre

nce

Det

ectio

n

Sev

erity

Rat

ing

Fac

tors

The FMEA Process

12


RPN = Risk Priority Number

General FMEA Process

Form Team, ID

Scope, Collect &

Review Data

Identify Function /

Requirements

Identify

Failure Modes

Identify Effects of

Failure Modes

Determine

Severity Rating

Identify

Possible Causes

Determine

Occurrence Rating

Calculate

Criticality

Identify

Detection

Controls

Determine

Detection Rating

Calculate RPN &

Set Priorities

Identify Special

Characteristics

Identify the

Root Cause

Recalculate RPN

& Determine

Next Actions

Take Actions to

Reduce Risk

start

end

Identify

Prevention

Controls

13


Software can speed up the FMEA process, but it is still important to keep the output

visible to all participants during a session.

Don’t let the typing in of information slow down the group process.

Consider using a combination of flip charts and sticky notes with the whole group, and

have someone type it into a spreadsheet later.

FMEA Team

• “Preparation” of the document—assign to an individual

• FMEA input, review and analysis is a TEAM effort

• Non-permanent team with representatives from:

– Process Engineering – Production Support

– Procurement – Service Delivery

– Quality – Production/Service

– Planning – Sales/Marketing/Program Management

– Finance – Customer Service/Support

– The CUSTOMER – Design & Development

• Best size: 4 to 6 people

• Goal: use FMEA as a tool to stimulate discussion & exchange of ideas

• Multiple FMEA Teams—ensure there is some overlap between teams

for coordination

14


FMEA Team

• Want cross-section of people:

✓ Some who are minimally knowledgeable – bring fresh perspective, more critical

✓ Some who are knowledgeable—bring understanding

• FMEA, by its nature, is critical —

“Let’s find everything that can go wrong” — some may take this

personally!

✓ Owner may tend to become defensive (“how dare you say my baby is ugly!”)

✓ May disregard or overlook possibilities (“that could never occur”)

✓ May lead to forming a Team that does not include owner of what is being

analyzed

15


A FMEA Team Start-up Worksheet is given in the Appendix.

FMEA Team

• Management needs to clearly define boundaries.

• FMEA team needs to understand boundaries—should not begin

without them!

• Answer following questions:

✓ What is Team responsible to do?

✓ What is the budget?

✓ What other resources are available or are not available?

✓ Is there a time constraint?

✓ How does the Team expand the boundaries (if need be)?

✓ What and how should the results be communicated? To whom?

• Should have a procedure for performing FMEA.

✓ Standing guidelines on boundaries

✓ Company’s defined process

16


See Appendix for examples of completed PFMEAs.

FMEA Worksheet

• Forms backbone of FMEA process.

• Can record actions using worksheet.

• Several versions — pick one you like or create your own (or

use one customer says to use).

• Worksheet should:

✓ Capture all important information/data

✓ Serve as a communication tool and record

✓ Be numbered (# should have some significance)

✓ Be changed as process design or plan progresses

✓ Be retained as a quality record (known location and retention

period)

• Typically — read and work with worksheet from left to right.

17

The ETI Group © 2020 18

Item

/Opera

tion A

rea:

PF

ME

A N

um

ber:

Level:

Page

of

Description:

Pre

pare

d b

y:

PF

ME

A D

ate

(O

rigin

al):

PF

ME

A D

ate

(R

evis

ed.)

:

Scope:

Core

Team

:

Pre

ve

ntio

n

Occurrence

De

tectio

n

Detection

Actions T

aken &

Co

mp

letion D

ate

Severity

Occurrence

Detection

CN

RP

N

Re

qu

ire

me

nts

To

tal R

PN

(b

efo

re a

ction):

Po

ten

tia

l

Ca

use

(s)

of

Failu

re

Severity

Po

ten

tia

l E

ffe

ct(

s)

of

Failu

re

Po

ten

tia

l

Failu

re M

ode

Cu

rre

nt

Pro

cess C

ontr

ols

Tota

l R

PN

(aft

er

action

):

Pro

cess S

tep

Action R

esults

Re

spo

nsib

ility

&

Ta

rget

Com

ple

tion

Date

Recom

me

nd

ed A

ctio

n(s

)R

PN

CN

Fun

ction

System

Process

Activity

Sub-System

PF

ME

A

Work

sheet

Process FMEA


Microwave Popcorn Exercise

1. We will create a high-level Flow Diagram of the process in the Large Group.

2. We will identify each Process Step’s Function and Requirements in the Large

Group.

3. Small groups will then be assigned to identify Potential Failure Modes and

Effects for one Process Step and linked Function-Requirements.

19


Each level relates to the others, especially for products manufactured with components nested

within assemblies, which are then nested within sub-systems, which are then grouped together to

form the system.

To use time most efficiently, create a hierarchy diagram as above to define context and

relationships before diving into the details of the FMEA. And, consider whether to focus “top

down” or “bottom up.”

• The top down approach to FMEA is from a process design standpoint, looking forward and

considering all possible failure modes in order to try to design them out or at least

minimize their effects. As said in the slide, beginning at a higher level will help prioritize

what to focus on in the process design/planning effort. Imagine if at the initial

design/planning of a system, an FMEA was performed for all activities in all processes in

all sub-systems…there could be an infinite number of possible failure modes identified!

The analysis process would quickly become overwhelming and unmanageable.

• When “debugging” from the bottom up, there are typically data on actual errors to work

from, either at the activity level or higher. What is learned here about process causes can

inform the design of the next generation of processes, or trigger changes to existing ones.

Both approaches are needed. If only actual failures were used in a FMEA, other potential failures

might get overlooked in the design/planning process.

Where to Focus?

System

Sub-systems

Assemblies/Processes

Components/Activities

Proactive Design Approach

Start at

System

level to

prioritize

where to

dig deeper

in lower

levels.

Reactive Debug Approach

Start at

Activity

level and

work

bottom-up

to address

design

issues.

20


Looking from the bottom up, the slide shows how a failure at a lower level may propagate

up through the various levels and ultimately impact the system, with an effect on the

customer.

The cause–failure mode–effect relationship between levels: the effects of failure modes

at lower levels become the failure modes at the next level up; the failure modes at lower

levels become the causes at the next level up.

As explained previously, the top down approach is used in a broader design sense. The

particular failure mode shown propagating up here could certainly be identified at the

beginning of a system level PFMEA. The difference being, it would likely be one of

several possibilities considered.

Relationships Between Levels

Cause Failure Mode Effect




System

Activities

Processes

Sub-systems

21


For any level, the above statement is true. For example, if a sub-system is chosen for

analysis, it will have emergent properties not present in the individual processes. The

interactions and interfaces of the sub-system with the environment/customers, of the

processes within the subsystem, the processes with the environment/customer and single

point failure will need to be considered (and so on down through the levels).

The idea is to set a practical boundary which will help in defining the Scope of the

FMEA. Depending on an organization’s role in the chain of design and/or manufacture

and the associated “field of view,” any operation could be considered a system, sub-

system, process or activity. Take for example a powder-coating process: it could be an

outsourced process in a higher level manufacturing flow, or the system itself if you are

the Powder Coater.

In a Process FMEA, the interfaces and interactions to be considered are between the “6

M’s” of the process: Machine, Method, Material, Measurement, Members, Mother Nature

(aka Environment). For non-manufacturing processes, some equivalents for the 6 M’s are:

equipment (e.g., computer/IS system) for machine, inputs/information for material,

feedback/reports for measurement, and work environment instead of physical

environmental factors.

Where to Start?

Each level has functions and relationships unique to that item as a whole

— its “emergent properties” — which are not present at lower levels.

Choosing a level (as opposed to jumping between levels in the same

FMEA), provides an opportunity to identify failure modes unique to the

interactions and interfaces that appear at that level alone.

System

Sub-systems

Environment &

Customers

A System level FMEA considers failure modes due to the interaction

and interfaces between:

✓ the system and its environment and customers;

✓ the sub-systems within the system;

✓ the sub-systems and the environment and customers,

✓ as well as single-point failures.

22


A FMEA can build on a previous one, especially for similar products, technologies, etc.

Design and Process FMEAs will also inform each other.

If the FMEA is in response to a change, the scope can just focus on the particular change.

For example, in the case of a machine or tooling modification, the FMEA could look at

possible interactions resulting from the modification. An FMEA could be performed to

assess the impact of a particular regulatory change; how a different operating

environment could affect a process; how using an existing process on a new product

could fail, etc.

The customer (and associated requirements) to be considered should follow from the

purpose of the FMEA. For example: End User/Distributor (reliability or on-time

delivery), OEM, Hub Sites or Supply Chain Assembly/Manufacturing Centers

(downstream manufacturability), Regulators (environmental and safety compliance),

Field Service (serviceability), etc.

It is also important to ask what is to be excluded from the Scope.

Diagramming the hierarchy (as shown previously) will be useful for initial scoping or

refining scope once a team is assigned. Other helpful tools are: DFMEAs for product to

be manufactured, Assembly Drawings, Process Flow Diagrams, Bills of Process (e.g.,

tooling and fixture lists), Interrelationship/Characteristic Matrices, etc.

FMEA Scope

The Scope sets the boundary of analysis.

A clear scope needs to be written and understood by the FMEA team

and Management.

It should answer the following questions:

✓ Is the focus design or debug (where is the product and/or process on the

development–to–launch timeline)?

✓ What operation or area is to be covered by this FMEA (system, sub-system, etc.)?

✓ What is triggering the FMEA (new product/process, changed environment, etc.)?

✓ Who is the customer and what are the customer requirements?

23


The PFMEA worksheet uses the term “process step.” Depending on scope, it could refer

to a step on a system level, sub-system, process or activity.

The format here should follow from what is currently used for process documentation in

order to link FMEA results to existing documentation. For example, an existing procedure

may have numbered process steps that correspond to a text description of the function, or

a flow diagram (aka process map) may show a process step with wording that adequately

describes the function (e.g., clean part in degreaser).

Process Step/Functions

• This work starts off the analysis portion of the FMEA worksheet and

is dependent on a clear Scope.

• List steps and interfaces with an amount of detail appropriate to the

level being analyzed.

• May need to create a more detailed flow diagram of operation being

analyzed.

✓ Can also indicate flow of information, product

✓ Use to understand inputs, function performed, outputs, interfaces

✓ Identify product/process characteristics and sources of variation associated

with each operation

✓ Walk the process!

Definition: Function is the intended purpose for a step which

relates to a product/customer need or expectation.

24


See Appendix for an example process flow diagram.

Identify Functions

• Give a simple description of the step and its purpose

✓ Apply surface coat to retard metal corrosion

✓ Tighten bolt using torque wrench

• If there are multiple functions for a step, list each separately

• Perform a risk assessment on the process

✓ Use knowledge of process, DFMEAs, past data to decide which steps need to be

analyzed

✓ Higher risk steps require PFMEA

25


Process Step Relative Risk

Get Job Packet Low Risk

Get tooling/fixtures Medium Risk

Install tooling/fixtures High Risk

Perform required Preventive Maintenance Medium Risk

Turn machine on, warm up Low Risk

Set parameters High Risk

Make set-up parts Medium Risk

Risk Assessment — Example for Equipment Set-up

Example Flow Diagram

Get Job Packet

Get

tooling/fixtures

Install

tooling/fixtures

Perform

required PM

Turn machine

on, warm up

Set parametersMake set-up

parts

Equipment Set-up

26


Process Step Relative Risk Requirement

Get Job Packet Low Risk —

Get tooling/fixtures Medium Risk Specified tooling/fixture for job

Install

tooling/fixtures

High Risk Correctly place and align

tooling/fixtures

Perform required

Preventive

Maintenance

Medium Risk Daily PM (coolant fill, lubrication)

Weekly/Monthly PM as required

Turn machine on,

warm up

Low Risk —

Set parameters High Risk Specified speed, pressure, feed rate

Make set-up parts Medium Risk Specified material per Bill of Material,

Make per Assembly Drawing

Requirements — Example for Equipment Set-up

Requirements

• List requirements for each function of the step being analyzed.

• If there are multiple requirements, write each one separately since they may have different failure modes.

Definition: Necessary inputs/outcomes of a step which relate to

product design intent and customer requirements.

27

Process FMEA

PFMEA Tree for Exercise #1

Item 1

Function 1

Function 2

Requirement 1

Requirement 2

Requirement 1

Requirement 2

Requirement 3


For system or sub-system level, the “steps” could be a few linked hand-offs between

functional groups and/or processes.

Project Application Exercise #1

1. Choose a Focus for your FMEA — either Design (looking forward for a

new process) or Debug (existing process with known issues, looking

backward at process design).

2. Choose a Level for your FMEA — System, Sub-system, Process or

Activity.

3. Name the operation you will analyze .

4. List a few steps in the operation along with their Functions.

5. For at least one of the steps, list Requirement(s) for each Function(s).

28


One assumption used is that the basic product design is correct, although the PFMEA may

identify design changes that could eliminate or reduce a failure mode (e.g., consideration

of tooling access, error proofing component placement, etc. ).

Another assumption is that incoming materials are correct, unless historical data proves

otherwise, and a consistent lack of capability must be considered.

Process StepRelative

RiskRequirement

Potential Failure

Mode

Get

tooling/fixtures

Medium Risk Specified tooling/

fixture for job

Wrong tool/fixture(s)

Install

tooling/fixtures

High Risk Correctly place and

align tooling/ fixtures

Incorrect Install

Set parameters High Risk Specified speed,

pressure, feed rate

Wrong parameters

Potential Failure Modes — Example for Equipment Set-up

Potential Failure Modes

Definition: The manner in which a step could potentially fail to

meet the requirements.

• If requirements are well-defined, the Failure Modes will follow easily.

• Use a technical statement: think in terms of an observable physical

event that could interfere with the process intent.

• For purposes of the analysis, assume that failure could occur

(“potential” means it won’t necessarily happen).

• Think about failure modes in terms of the Scope.

• List potential failure modes of each function/requirement.

29


A large number of failure modes for a single function/requirement may be another clue

that better definition is needed.

Brainstorming and Affinity Diagrams can be useful tools for coming up with failure

modes.

Afterward, it can be helpful to review historical information for the operation (or a similar

one) to validate the identified failure modes. Info can include past quality & reliability

reports, field service reports, process control reports, test results, customer complaints,

warranty data, etc.

Potential Failure Modes

• Types of failure mode:

✓ Complete failure

✓ Partial failure

✓ Intermittent failure

✓ Failure over time

✓ Over-performance of function

• Categories help us avoid overlooking potential failure modes —

consider each type.

• May also help point out if Functions are poorly defined.

30


Be specific—operator fails to connect tab vs. operator error; bearing fails vs. machine

malfunction.

Examples of descriptors for failure modes:

– Bent – Cracked – Tool Worn – Open circuit

– Short circuited – Grounded – Handling Damage – Binding

– Improper Set-up – Deformed – Burred – Dirty

Let’s look at how a Failure Mode identified for the Equipment Set-up example could be

expanded to include these 5 Failure Mode categories:

Failure modes—Example

Initial Function Definition:

“Apply surface coat to retard metal corrosion.”

Failure Mode categories:

Complete: No coating

Partial: Insufficient coating coverage

Intermittent: One part has coating, next doesn’t

Failure over time: Gradual decrease/increase in coating thickness

Over-performance: Too much coating

Poorly defined function:

✓ If there is too much coating, it over-performs but still meets the initial Function definition.

✓ Change definition to “Cover surface with minimum coating thickness to retard corrosion.”

31

Process

StepRequirement

Initial Potential

Failure Mode

Expanded Potential Failure

Modes

Install

tooling/

fixtures

Correctly place

and align

tooling/ fixtures

Incorrect Install Complete:

Partial:

Intermittent:

Failure over time:

Over-performance:

Potential Failure Modes — Example for Equipment Set-up



1. For the step you chose in Exercise #1, pick one of your identified

function/requirements to use for this exercise.

2. Identify Potential Failure Modes of the function/requirement;

consider all 5 types of Failure Modes:

• Complete

• Partial

• Intermittent

• Failure over time

• Over-performance

32


Item 1

Function 1

Function 2

Requirement 1

Requirement 2

Requirement 1

Requirement 2

Requirement 3

Failure Mode 1

Failure Mode 2

Failure Mode 3

Failure Mode 4

Failure Mode 5

Complete

Partial

Intermittent

Failure Over Time

Over-performance


There can be multiple effects for each potential failure mode; list all effects and

consequences.

Potential Effects & Consequences

Definition: Effects of the failure mode as perceived by the

customer.

• Non-technical statements that describe symptoms of the failure mode.

• List effects (and consequences) for each failure mode.

• Consider various levels of customers leading up to the End-user.

• Regardless of design specifications, consider what would be

objectionable to the customer.

• Identify any safety or regulatory/legal impacts

33

Process Step Failure Mode Effect Consequence to Customer

Apply surface

coat

Gaps in surface

coat

Unsatisfactory surface

finish

Mfg: If detected, sorting & reworking needed

End-user: Parts rejected or if used, later corrosion

damage risk

Tighten bolt Bolt too loose Mechanical instability

(part can swing on frame)

Mfg: Difficult to attach next part

End-user: Parts can vibrate, will loosen over time

and come apart causing product failure.


FM = Failure Mode

The description of effects for the operation under analysis and any next operations

(downstream assembly in-house or at other locations) should be stated in terms of process

operation and performance.

End-user effects should be stated in terms of product effects, i.e., the consequences

experienced with the product.

What might be some effects and consequences for the previously identified Failure

Modes for the Equipment Set-up example?

A Flow Chart for Effects Determination

Consider Effects in this order:

1. Operation under analysis

2. Next operation(s)/location(s)

3. End-user

List effects of FM if it

occurs &/or is

detected

Does

occurrence of

FM physically

prevent further

processing?

Stop here

Yes

No

34

Process StepPotential Failure

ModeEffects & Consequences

Get tooling/fixtures Wrong tool/fixture(s)

Install

tooling/fixtures

Incorrect Install

Set parameters Wrong parameters

Effects & Consequences — Example for Equipment Set-up


OR Gate: events 2 AND 3 OR event 4 will occur as a result of event 1.

AND Gate: events 2 AND 3 will happen together as a result of event 1.

Related Tool: Event Tree Analysis

• Visual representation of all the

events that can occur in a system.

• Starts at a root event and follows

what happens because of that

event.

• Good tool to trace linking events to

see consequences.

• Goal is to determine probability of

a future event based on outcomes

of prior events in the chronological

sequence leading up to the event

being considered.

Time

1

2

3

4

OR

AND

35


A five point scale may also be used.

Always “round” to higher value for rating numbers.

Any information on past nonconformities and lessons learned, data on quality, reliability,

capability, customer complaints, warranty returns, etc. will be helpful in determining

ratings. A survey of information available on best practices (such as guidelines and

standards for manufacturing, identification and traceability, error proofing, etc.) will also

help.

Assigning Ratings

• Severity, Occurrence & Detection all based on 10 point scale.

• 1 is best case, 10 is worst case.

• Need clear, concise definitions for rating scale—as descriptive as

possible.

• Establish before beginning rating process.

• Organization can customize rating scales or use industry standard

scales:

✓ Best to have company standards

✓ Scales should be specific to product or process

36


Severity ratings should be considered as appropriate for the various customer effects, both

for in-house operations and the End-user. The final Severity rating will be the overall

highest number.

Severity Rating

• First step in analyzing risk is to rank severity of effect and resulting

consequences to the customer.

• Assume that effects will result given that failure mode has occurred

• Effects/consequences are rated as a group—most serious consequence

establishes the rating (“worst case”)

• Use experience, knowledge, expertise to decide.

• Rank is relative within the scope of the FMEA

Definition: Value assigned to the most serious effect for a given

failure mode.

37


When a portion of the product has to be reworked or scrapped, this action also implies

sorting.

Severity Rating—Manufacturing

Rating Description Consequence

10

Hazardous – no

warning

Failure affects safe process operation (may endanger

personnel or equipment) or involves noncompliance with

government regulation. No warning of failure provided.

9Hazardous – with

warning

Ditto above except failure occurs with a warning.

8Very High 100% of product may have to be scrapped. Line or shipment

stoppage.

7

High Portion of product may have to be scrapped. Deviation from

primary process needed (change to process

parameters/conditions, decreased production rate, additional

resources required).

6Moderate 100% of product may have to be reworked off-line and

reverified.

5 Low Ditto above except only a portion of product is affected.

4Very Low 100% of product may have to be reworked in-station before

continuing processing.

3 Minor Ditto above except only a portion of product is affected.

2 Very Minor Slight inconvenience during process operation.

1 None No discernible effect.

5

4

3

2

1

Failure to meet

Safety &/or

Regulatory

Requirements

Major Disruption

Moderate Disruption

Minor Disruption

No Effect

Significant Disruption

Slight Disruption

38


Severity Rating—End-user

Rating Description Consequence

10

Hazardous – no

warning

Failure affects safe product operation or involves

noncompliance with government regulation. No warning

provided when failure occurs.

9

Hazardous – with

warning

Failure affects safe product operation or involves

noncompliance with government regulation. Failure occurs

with a warning.

8 Very High Product is inoperable with loss of Primary Function.

7 High Product is operable but at a reduced level of performance.

6 Moderate Product is operable but convenience item(s) are inoperable.

5Low Product is operable but convenience item(s) operate at a

reduced level of performance.

4Very Low Fit and finish or squeak and rattle item does not conform.

Most customers notice defect.

3Minor Fit and finish or squeak and rattle item does not conform.

Average customers notice defect.

2Very Minor Fit and finish or squeak and rattle item does not conform.

Discriminating customers notice defect.

1 None No discernible effect.

5

4

3

2

1

Failure to meet

Safety &/or

Regulatory

Requirements

Loss or

Degradation of

Primary Function

Loss or

Degradation of

Secondary

Function

Annoyance

No Effect

39

For service delivery applications, the “End User” rating scale is most appropriate; just

think “service” for product and “deliverable” for “operable.”



1. For each of the failure modes identified in Exercise #2, list the Effects.

Consider effects for each of the different levels of customers leading up to

the End-user.

2. For each linked Failure Mode and Effects chain (do at least 3), give a

Severity Rating for both the manufacturing and End-user Effects.

3. Assign an overall Severity rating for each linked Failure Mode and Effects

chain.

4. Use the Severity rating scales provided on the previous pages.

40


Item 1

Function 1

Function 2

Requirement 1

Requirement 2

Requirement 1

Requirement 2

Requirement 3

Failure Mode 1

Failure Mode 2

Failure Mode 3

Failure Mode 4

Failure Mode 5

Effects

Severity

Severity

Severity

Severity

Severity

EffectsEffects

Effects

Effects

Effects

EffectsEffects

Effects

Effects

Effects


Initially, focus on the immediate cause of the failure. The purpose at this point is to be

able to determine Occurrence rating.

Deeper root cause analysis will be performed later on high priority causes (during the

“Recommended Action” step).

The same assumptions regarding product design and incoming material/components used

when determining failure modes apply here.

Potential Cause

• Real world failure modes often have multiple potential causes; list

each one separately.

• Causes are potential — may or may not occur.

• Describe in terms that allow correction or control.

• Consider variation sources (remember the 6 M’s).

• Begin with Failure Modes that have the highest Severity rating.

Definition: The condition that makes it possible for a failure

mode to occur.

41


Avoid generic, ambiguous phrases like “poor processing” or “operator error.”

See Appendix for an example Cause and Effect Diagram

What might be some potential causes for the previously identified Failure Modes for the

Equipment Set-up example?

Potential Cause

• Be as concise and complete as possible.

• Good tools to use – Cause & Effect Diagram or Fault Tree Analysis.

• Example Causes:

✓ Improper torque—over or under

✓ Improper weld—current, time, pressure

✓ Inaccurate gauging

✓ Improper heat treat—time, temperature

✓ Inadequate gating or venting

✓ Inadequate or no lubrication

✓ Part missing or wrong location

✓ Wrong tooling/fixture

42

Process StepPotential Failure

ModePotential Causes

Get tooling/fixtures Wrong tool/fixture(s)

Install

tooling/fixtures

Incorrect Install

Set parameters Wrong parameters

Potential Causes — Example for Equipment Set-up


AND Gate: events 2 AND 3 AND 4 have to happen for event 1 to occur.

OR Gate: event 5 OR 6 has to happen in order for event 4 to occur.

See Appendix for more details on FTA

Related Tool: Fault Tree Analysis

• Deductive, top-down method.

• Analyzes system design and

performance.

• Specify top event to analyze and

determine everything that would

have to happen for event to

occur.

• Makes use of logical AND and

OR gates.

• Should analyze entire system as

well as human interactions.

3 4

Time

1

5 6

2

AND

OR

43


Project Application Exercise #4-a

1. Choose a failure mode with a high Severity rating from Exercise #3.

2. Brainstorm potential Cause(s) of the Failure Mode(s).

44

PFMEA Tree for Exercise #4-a

Item 1 Func. 1 Req. 1

FM 1

FM 2

FM 3

FM 4

FM 5

E’s S

E’s S

E’s S

E’s S

E’s S

Cause

Cause

Cause


SPC can be considered a prevention control for a particular cause only if it is used real-

time and can predict the occurrence of the cause before nonconforming parts are made

(e.g., tool wear). This ability requires a thorough understanding of how process variation

relates to product characteristics.

Current Process Controls — Prevention

• The occurrence rating will be affected by the preventive process

controls.

• “Current” means have been used on same or similar process.

• Some Prevention Process Control Examples:

✓ Statistical process control (SPC) charts

✓ Error proofing of tooling/fixtures

✓ Error proofing of data entry

✓ Set-up sheets

✓ Preventive and predictive maintenance

✓ Operator training/qualification

✓ Built-in equipment controls/protection features

Definition: Process–level controls that eliminate (or reduce the

rate of) the occurrence of a cause of a failure mode.

45


Occurrence Rating

• Occurrence rating is relative within the scope of the FMEA; it should

not be confused with actual likelihood.

• Do not evaluate causes as a group—each cause gets its own occurrence

rating.

• Consider:

✓ Failure rates, service history, field experience with similar process

✓ Whether item is a carryover from or similar to a previous process

✓ If operation is new, significantly changed/different from previous process

✓ Changes to application/operating environment, etc.

✓ Results of reliability or other Engineering analysis

Definition: Likelihood that a particular cause will occur and result

in the failure mode.

46


Note: A Cpk value for a defect rate of zero is mathematically impossible, but for reference,

a Cpk = 2 will give an estimate of 1 part per billion defective.

Occurrence Rating

Rating DescriptionFailure

RateCpk

Likelihood

of Failure

10New technology/process with no history; failure is

inevitable.

> 1 in 10 > 0.43 Very High

9Significant failures associated with similar processes. 1 in 20 > 0.55

8 Regular failures associated with similar processes. 1 in 50 > 0.69 High

7 1 in 100 > 0.78

6 Frequent failures associated with similar processes. 1 in 500 > 0.96 Moderate

5 Occasional failures associated with similar processes. 1 in 2,000 > 1.1

4 Isolated failures associated with similar processes. 1 in 10,000 > 1.24

3Only isolated failures associated with almost identical

processes.

1 in 100,000 > 1.42 Low

2No observed failures associated with almost identical

processes.

1 in

1,000,000

> 1.58 Very Low

1Failure is eliminated through preventive process control. cannot

happen

NA Remote

5

4

3

2

1

47


Project Application Exercise #4-b

1. Use the potential Cause(s) identified in Exercise #4-a.

2. List Preventive Process Controls for each potential Cause.

3. Give each Cause a likelihood of Occurrence using the rating scale

provided on the previous page.

48

PFMEA Tree for Exercise #4-b

Item 1 Func. 1 Req. 1

FM 1

FM 2

FM 3

FM 4

FM 5

E’s S

E’s S

E’s S

E’s S

E’s S

Cause Prevtn. Controls Occurrence




Special Characteristic Definition:

Product characteristic or operational process parameter which can affect safety or

compliance with regulations, fit, function, performance or subsequent processing of

product.

Implementation of Special Characteristics:

• Could be addressed as part of Risk Management activities

• May get identified with a designated symbol on drawings (e.g., functionally

critical, automation/assembly critical, etc.)

• Could be added to plans for process control and test/inspection

General rule of thumb:

• Values of 9 & 10 (highest numbers) must be addressed

• Value of 1 should result in no further action

Identifying Special Characteristics

10987654321

1

2

3

4

5

6

7

8

9

10 CRITICAL CHARACTERISTICSSafety / Regulatory

SIGNIFICANT

CHARACTERISTICSCustomer Dissatisfaction

ALL OTHER CHARACTERISTICSAppropriate actions / controls already in place

S

E

V

E

R

I

T

Y

OCCURRENCE49


Note: If using a 1 to 5 scale, first multiply ratings by 2.


1. Plot the Severity and Occurrence Ratings from previous exercises on

the Severity x Occurrence Chart on the “Identifying Special

Characteristics” slide.

2. Would your plots lead to the identification of any Critical &/or

Significant Characteristics?

50


Detection controls may happen at an operation or any time before product is sent to an

external (paying) customer.

SPC cannot be considered a detection control because it uses process sampling and is not

intended for sorting nonconforming product.

Current Process Controls — Detection

Definition: Controls that detect the cause of failure or the failure

mode after they have happened and which lead to

controls for nonconforming product.

• Detection controls affect the Detection rating.

• “Current” means have been used on same or similar process.

• Some Detection Process Control Examples:

✓ Inspection/test at or downstream from operation

✓ Final product acceptance inspection/testing

✓ Measurement validation at set-up (torque, pressure, temperature, etc.)

✓ On-line go/no-go sensors

✓ Automated on-line visual inspection

51


Detection Rating

• Detection rating is relative within the scope of the FMEA, it is not a

guarantee of performance.

• Assume that the failure mode has taken place and assess the ability of

the detection process controls to find it.

• Do not evaluate causes as a group—each cause gets its own detection

rating.

• If there are multiple detection process controls for a cause, list a rank

next to each but record only the lowest number (best control method) in

the Detection column.

• Do not presume that detection rank is low if occurrence rank is low,

these are two different things!

Definition: Rank given to the best current detection process

control.

52


FM = Failure Mode

Opr = Operator

“Visual/Audible/Tactile” (v/a/t) techniques could be an operator looking at parts as they

come off a machine, running a hand over a surface to check for burrs, listening for a

correct/incorrect sound or other “human senses” checks.

Detection Rating

Rating DescriptionLikelihood of

Detection

10 No current process control; cannot detect cause/failure mode or is not analyzed. Almost Impossible

9 Cause/FM is not easily detected (e.g., random audits of process conducted). Very Remote

8FM checked for by opr post-processing through visual/tactile/audible (v/a/t) technique. Remote

7FM checked for in-station by opr through v/a/t technique or post-processing through

use of attribute gauging (go/no-go, manual torque check/clicker wrench, etc.).

Very Low

6FM checked for post-processing by opr through use of variable gauging or in-station by

opr through use of attribute gauging (see examples above.).

Low

5

FM/Cause checked for in-station by opr through use of variable gauging or by

automated controls in-station that will detect N/C part & notify opr (light, buzzer, etc.).

Set-up causes only: Gauging performed on set-up & first-piece check.

Moderate

4FM checked for post-processing by automated controls that will detect N/C part & lock

part to prevent further processing.

Moderately High

3FM checked for in-station by automated controls that will detect N/C part &

automatically lock part in station to prevent further processing.

High

2Cause checked for in-station by automated controls that will detect error & prevent N/C

part from being made.

Very High

1Cause prevented due to design of fixture/tooling/machine/part. N/C parts cannot be

made because of error-proofing of process/product design.

Almost Certain

5

4

3

2

1

53



1. For the same cause(s) identified in Exercise #4, list Process Controls that

would Detect each failure mode or cause(s).

2. Rate each Detection Process Control with a likelihood of its ability to detect

the failure mode or cause(s).

3. Use the Detection rating scale provided on the previous page.

4. If there is more than one Detection Control per cause, list each, but use

lowest (best) ranking for the final Detection number.

54


Item 1 Func. 1Req. 1

FM 1

FM 2

FM 3

FM 4

FM 5

E S

E S

E S

E S

E S

C Prevtn. O Detectn. Controls Detection




RPN is a relative measure of risk

Like a golf score, lower is better

Risk Priority Number (RPN)

RPN = Severity x Occurrence x Detection

55


It is important to remember that the RPN is:

• A relative number

• Not a calculated reliability metric

• Not continuous data

RPN Cautions

• RPN is distorted because of non-linear scales in different

individual scales.

• Severity, Occurrence and Detection are weighted equally.

• Incorrect assumptions:

Incorrect Assumption Actual Statistical Data

The average of all RPN values is roughly 500 The Average RPN value is 166

Roughly 50% of RPN values are above 500

(the median is near 500)

6% of all RPN values are above 500

(the median is 105)

There are 1000 possible RPN values There are 120 unique RPN values

56


The Criticality Number is what gets plotted on the “Special Characteristics” chart shown

earlier.

RPN Conclusions

• RPN should be used as a guideline to rank concerns, aid dialog

and prioritize action.

• High Severity values merit special attention, especially if

coupled with high Occurrence values.

• The Criticality Number (CN = S X O) is an additional tool to

use.

57


While it may seem helpful to establish a “cutoff RPN,” there is a risk of driving the

wrong behavior and not taking action when it is merited.

As mentioned before, RPN and CN are helpful in ranking priorities. Action should then

proceed on the basis of which risks are most important to mitigate, given the time and

resources available.

RPN vs. CN — Example

• RPN, alone, can be misleading.

• Should calculate both CN and RPN and use together for ranking

issues.

• Use of a threshold or cutoff value for action is not recommended.

Severity Occurrence Detection CN RPN

8 6 1 48 48

1 6 8 6 48

RPN = S x O x D

CN = S x O

58


Which failure modes account for 80% of the total RPN?

Raw Data for Pareto Chart—Equipment Set-up Failure Modes

Prioritizing Failure Modes — Example

RPN/CN Pareto chart for Equipment Set-up

37%

27%

14%12%

8%

3%29% 21%

7% 13%21%

8%

0

50

100

150

200

250

300

350

400

Misaligned

Fixture

Wrong Pressure

Setting

Wrong Material

Pulled

Wrong Tooling

Used

Missed Lube

Check

Wrong Rev Dwg

RPN

CN

59

Category Severity Occurrence Detection RPN % RPN CN % CN

Misaligned Fixture8 6 7 336 37% 48 29%

Wrong Pressure

Setting 8 5 7 280 27% 40 21%

Wrong Material

Pulled2 7 10 140 14% 14 7%

Wrong Tooling

Used5 5 5 125 12% 25 13%

Missed Lube

Check10 4 2 80 8% 40 21%

Wrong Rev Dwg3 5 2 30 3% 15 8%

Total RPN 991 182


Which failure modes would be selected if the top 80% were used?

How do the top three failure modes compare between RPN and CN?

Raw Data for Pareto Chart—Equipment Set-up Failure Modes

Category Severity Occurrence Detection CN % CN

Misaligned Fixture 8 6 7 48 29%

Wrong Pressure Setting 8 5 7 40 21%

Missed Lube Check 10 4 2 40 21%

Wrong Tooling Used 5 5 5 25 13%

Wrong Rev Dwg 3 5 2 15 8%

Wrong Material Pulled 2 7 10 14 7%

Total RPN 182

Prioritizing Failure Modes

Pareto chart of CN Only for Equipment Set-up

29%

21% 21%

13%

8% 7%

0

10

20

30

40

50

60

Misaligned Fixture Wrong Pressure

Setting

Missed Lube

Check

Wrong Tooling

Used

Wrong Rev Dwg Wrong Material

Pulled

60



1. Calculate CN and RPN for your Ratings.

2. Calculate total CN and RPN (and sketch Pareto charts as desired).

3. What would your priorities be?

61


Item 1 Func. 1Req. 1

FM 1

FM 2

FM 3

FM 4

FM 5

E S

E S

E S

E S

E S

D RPN CN

C Prevtn. O Detctn. D RPN CN

C Prevtn. O Detctn.

C Prevtn. O Detctn. D RPN CN


This is the time to dig deeper into the causes and the reasons they occur. Earlier work on

cause determination can be revisited at this point, especially if Cause and Effect

Diagrams, Fault Tree Analysis or other tools were used.

The “Five Why’s” is another helpful technique for getting to root causes. Brainstorming,

Affinity diagrams and data collection/analysis can also be used both for investigating

causes and generating ideas for preventing their occurrence.

If no recommended action is taken on a particular failure mode/cause/control chain, it

may be helpful to future readers (and prudent in the case of external customers/regulators)

to give the rationale (especially for steps with a high severity rating).

Reducing Risk

• Fundamental purpose of performing FMEAs is to

recommend and take actions that reduce risk.

• Direct actions at critical items and highest ranked items.

• Can reduce risk by lowering Severity, Occurrence, or

Detection rating.

• Decisions & action taken should be noted on FMEA

worksheet.

62


Severity can often be difficult to impact, especially so once a product design has been

released or a process has been implemented. Making a process/product design revision

does not automatically mean severity will be reduced. For best success with reducing

severity, FMEAs should be performed early in the development process.

Improving and/or increasing detection controls (e.g., inspection/test) is generally costly

and not as effective as preventive controls, even though this technique is very typical in

many organizations. Increased inspection/test could be used temporarily to gather

sufficient data to study significant factors in process variation (but a Designed

Experiment would be better!).

Improving employee and customer (downstream, end user) training may impact all three

rankings.

Possible Actions to Reduce Ratings

Severity Occurrence Detection

• Design Revision to:

✓ Improve design of

process/product

✓ Compensate or mitigate

severity of failure

• Change application &/or field

environment

• Design Revision to remove or

control cause:

✓ Error proof product/process

✓ Revise design geometry &

tolerances

✓ Fail safe process design

✓ Increase Cpk through Design

of Experiments (DOE) &/or

equipment modification

• Improve knowledge of dominant

causes of process variation

(special & common) &

interactions (DOE, reliability

testing)

• Ensure the measuring devices

are accurate and precise (regular

calibration and gauge studies)

• Ensure Preventive Maintenance

performed

• Parts/component change to make

detection easier

Reduce rankings in this order

63


Example Actions

Reduce rankings in this order

Severity Occurrence Detection

Process design change:

• Personal protective equipment

(e.g., hard hats, side shields on

safety glasses, full face

protection, ventilators, etc.)

• Safety stops or emergency

shut-off’s

• Use of different materials (e.g.,

safety glass, chemical

alternatives) that will not cause as

severe an injury

Product design change:

• Choose product

components/materials that won’t

be damaged by processing

conditions

• Engaging mechanism that must

be activated for machine to work

(e.g., lawnmowers with handle

that must be held to run)

• Automated feedback loop

between a product characteristic

and process parameter

• Total Predictive Maintenance

(TPM) to replace tooling, bearing,

etc. before Production parts

become nonconforming

• See also section on Prevention

Process Controls for ideas

• Use coding (colors, shapes,

sounds) to alert the operator that

something is either right or wrong

• See also section on Detection

Process Controls & the Detection

Rating Scale for ideas

64



1. If resources were not an issue, what would be some recommended

actions for the priorities identified in Exercise #7?

2. Imagine that these actions were taken. Revisit the Severity,

Occurrence and Detection Ratings and revise them as appropriate.

3. Recalculate CN and RPN.

65


Following Up

RPN should be recalculated after recommended actions are

taken and proven to be effective.

Evaluation could include:

• Ensuring that Process Capability goals are achieved.

• Confirmation studies (DOE, reliability) to confirm solution and

assure changes have not introduced new failure modes.

• Reviewing changes made to floor layout,

machine/tooling/fixture specifications, bills of process,

preventive maintenance plans, etc. to confirm changes.

• Confirming changes to standards, guidelines, procedures

(including process flow diagrams, work instructions, etc.),

process control plans and other documentation.

66


Think of FMEA as an iterative continuous improvement tool.

See the Appendix for more information on quality planning.

FMEA Management

Use Process FMEA recommended actions for ongoing control:

• Identify special characteristics and parameters.

• Ensure prevention and detection controls continue to be used.

• Develop process control and inspection/test plans.

• Provide feedback and ideas for product design

Periodically review FMEA:

• Validate/revise estimates through post launch data (internal and

external).

• Pay special attention to occurrence and detection ratings.

• Revisit FMEA whenever a process change occurs.

• Maintain revision control.

67


Summary of Process FMEA

• Goals:✓ Identify potential process failures

✓ Assess customer effects of failure

✓ Identify process causes and variables to focus controls

✓ Reduce severity and/or occurrence or increase detection

✓ Develop prioritized list of issues to correct/improve

✓ Record thoughts, decisions and plans as product is designed

• Should include all operational and support processes

• “Customer” is End User and down-stream operations

• Parallels, formalizes and documents engineering mental disciplines

• Perform on all new processes, changed processes, and processes used in new

applications or environments

• Assumes product, as designed, will meet the design intent

• Does not rely on product design changes to overcome process weaknesses (but

does consider product’s characteristics)

68


Key Elements of FMEA

Performed early and continuously throughout the

development cycle to reduce risk of failure and

decrease product/process design changes late in the

development cycle.

Conducted by a team of people with diverse

experience and expertise (including customer

representation if possible).

Provides a structured “living” document to be

updated throughout the design and development

process — records the basis for design decisions

and process controls. An excellent tool for

continuous improvement.

Timeliness

Knowledge

Teamwork

69


This page is intentionally blank.


Appendix


FMEA Number: Date Started:

Date Completed:

Team Members:

Team Leader:

1. Are all affected areas represented on the team?

YES NO Action:

2. Are different levels and types of knowledge represented on the team?

YES NO Action:

3. Is the customer involved?

YES NO Action:

4. Who will take minutes and maintain records?

FMEA Team Boundaries of Freedom

5. What aspects of the FMEA is this team responsible for?

FMEA Recommendations for Implementation of Analysis Improvement Improvements

6. What is the budget for the FMEA?

7. Does the project have a deadline (provide if Yes)?

8. Do team members have specific time constraints?

9. What is the process to follow if the team needs to expand beyond these boundaries?

10. How should the FMEA results be communicated to others?

11. What is the scope of this FMEA? (be specific and include a clear definition of the process or product to be studied)

FMEA Team Start-up Sheet

Process FMEA


Process Flow Diagram — Example

Open Package

Screen Solder Paste on

Board

Remove Part From Package

MANUAL LOAD

Place Part on Board

Place Package on Line

Screen Solder Paste on

Board

Pick Product From Package

Vision Inspection for

Location

AUTO LOAD

Place Product on Board

IR Solder Reflow

Clean Board

Inspect/Test

Assembly to Mating Board

Rework

Pass

Fail

Customer process for board-to-board SMT connector

73


This diagram is an example of how an organization used the potential causes of a failure

mode determined during an FMEA session to create a trouble-shooting guide. It is not a

“classic” C & E Diagram because while it has a typical “effect” — in this case “poor

solder connection to a printed circuit board” — it does not use the typical “6M” sources

of process variation (Machine, Method, Material, Measurement, Members, Mother

Nature/Environment). For the organization, it was a format that operators were already

familiar with, so they chose to show it using the “bones” of the diagram for different

cause categories. Any number of formats could be used; what’s important is to consider

how the results of the FMEA work can be used to assist in ongoing process control

efforts.

Cause-and-Effect (Fishbone) Diagram — Example

Poor Solder

Connection

to PCB

Poor Solder Ball Integrity

Size

Volume

Oxidation

Solder

Connection

Geometry

Mis-Located on PCB

Low Strength

Solder Joint

Hsg. Warp

Z Axis Cap

Ball Size Variation

Pad Size

Shape

Solder Ball

PositionFractured

Joints

Packaging

Gripper Datum on Hsg., XYZ

Orientation

Missing

X-Y Pos.

Hsg.

BallContact Tip Position

Copper/Mask

Materials

Poor

Matl.

Selection

CTE

Plastic Inclusions

Hsg. Lead-In

Geometry

Retention Gap

Too SmallContact

Insertion

Tip Geometry

Housing Flash

Voids

Intermetalics

Overheating

Gold Embrittlement

Contamination

Incomplete Reflow

Warp

Board/Connector

Temp/Time

Customer

Plastic Inclusions

Poor solder connection to PCB

74

Process FMEA

AND Gate: events 2 AND 3 AND 4 have to happen for event 1 to occur.

OR Gate: event 5 OR 6 has to happen in order for event 4 to occur.

Method is deductive, top down, where #1 above is the Failure Mode and subsequent

numbers are possible causes.

Logical AND and OR gates are used to show everything that would have to happen for

the failure mode to occur.

A systems view should be considered and include human interactions (think in terms of

the 6 M’s).


Fault Tree Analysis

75

• Fault Tree Analysis originated with Bell Labs

• Element of Failure Modes & Effects Analysis

✓ Used to analyze all possible causes of a particular failure mode

• In practice, it looks like a combination of tools

✓ Kepner Tregoe

✓ Cause & Effect Diagrams

✓ The 5 Whys

• Powerful graphical tool that is easy

and intuitive to use


Where Do You Start?

76

Start by asking “What could cause this Failure Mode?”

• Think about causes that could occur separately (OR

gates)

• Consider combinations of errors/failures that would

occur together (AND gates)

• May need to “walk the process”

• Use historical data and experience

• Lay out the Fault Tree Diagram showing the possible

causes in sequence to each other, starting at the top level

and drilling down (sticky notes are helpful)


Example: Fault Tree — 1st Level Analysis

77

• What’s The Problem?

Lathe turned part has poor surface finish in ID of wall.

• Possible Causes?

Wrong Material Specified (Design FMEA)

Wrong Tool (Process FMEA)

Tool Chatter (Process FMEA)

Process FMEA

A table of possible combinations may be a helpful aid to discussion, to determine whether

an “OR” and “AND” situation is present. A “T” means the condition is “true,” i.e.,

present. “F” means it is not present. As shown above, if situation 1, 2, or 3 is present, the

Failure Mode will be present. The causes are not dependent on each other, therefore an

“OR” gate is appropriate.


Fault Tree — 1st Level Layout

78

Surface Finish Too Rough

Wrong Material Wrong Tool

Tool Chatter

1 2 3 Fail?

T T T T

F F F F

T T F T

F F T T

T F F T

F T F T

T F T T

F T T T

Or Gate

1 2 3


Fault Tree — 2nd Level Analysis

79

• Wrong material specified

Incorrect Alloy

Incorrect Temper

• Wrong Tool

• Tool Chatter

Wrong Feed Rate

Machine Problem1.11R 1.38R


Fault Tree — 2nd Level Layout

80

Surface Finish Too Rough

Wrong Material Wrong Tool

Tool ChatterIncorrect Alloy

Incorrect Temper Wrong Feed Rate

Machine Problem

Process FMEA

Here, the table for the “AND” gate shows the dependency between the two sub-causes for

Incorrect Temper — both have to be true in order to result in a failure.

This example just shows one path. The analysis could continue to drill down further into

causes for the path shown and the other causes.


Fault Tree — 3rd Level for Wrong Material (DFMEA)

Not detected at

Design Review

Design Guideline not

followed

1 2 Fail?

T F F

F T F

T T T

F F F

And Gate

1 2

81

Process FMEA

Here, the table for the “AND” gate shows the dependency between the two sub-causes for

Wrong Tool— both have to be true in order to result in a failure.

This example just shows one path. The analysis could continue to drill down further into

causes for the path shown and the other causes.


Fault Tree — 3rd Level for Wrong Tool (PFMEA)

82

1 2 Fail?

T F F

F T F

T T T

F F F

And Gate

1 2

Wrong tool not

detected

Wrong tool PN

on set-up sheet


Take advantage of the Rule of 10’s!

Quality Planning with FMEA

• Product & process planning and risk management:

✓ Should follow a structured method

✓ Define and establish the steps needed to assure product satisfies the

customers’ requirements

✓ Should be an integrated system – design leading into process planning

✓ Can use analytical techniques such as DOE, Value Engineering, simulation,

Computer Aided Engineering (CAE), etc.

• Process control plan:

✓ Aids in producing product that meets requirements

✓ Goal is to minimize variation

✓ Structured approach for design, selection & implementation of value-added

control methods for total system

✓ Use a team approach with people experienced on similar product/process

• Design & Process FMEAs are important tools to use as part of the

planning process.

83


Manufa

ctu

ring P

rocess F

ME

A —

Exam

ple

for

Equip

ment

Set-

up

Process FMEA

Item

/Opera

tion A

rea:

Equ

ipm

ent

Set

-up

PF

ME

A N

um

ber:

Leve

l:

Page

1of

1

Description:

S

et-u

p o

f th

e b

ifur

cula

tor

mac

hine

use

d o

n th

e d

imid

iato

r pro

duc

t.P

repare

d b

y:

PF

ME

A D

ate

(O

rigin

al):

PF

ME

A D

ate

(R

evi

sed.)

:

Scope:

W

ill an

alyz

e pro

cess

fro

m w

hen

a jo

b p

ack

et is

pul

led

(al

thou

gh t

he j

ob p

ack

et c

reat

ion

pro

cess

will b

e co

vere

d b

y a

dif

fere

nt P

FM

EA

)

thro

ugh t

he p

rod

ucti

on o

f se

t-up

par

ts. S

et-u

p p

art

insp

ecti

on is

excl

uded

fro

m t

his

scop

e an

d w

ill b

e per

form

ed a

s a

separ

ate

PFM

EA

.

Core

Team

:

B.E

. Kin

de

(QA

), T

.U. A

ahl

(M

fg. E

ngr.

), Y

.R. F

reun

dz

(Mai

nt.)

, E

.V. R

idai

(M

fg. L

ead

)

Pre

vention

Occurrence

Dete

ction

Detection

Actions T

aken &

Com

ple

tion D

ate

Severity

Occurrence

Detection

CN

RP

N

Low

Ris

k1.

G

et

Job

Packet

(not

ad

dre

ssed

in t

his

PF

ME

A)

Med

ium

Ris

k

2. G

et

tooli

ng &

fix

ture

s

Specif

ied

tooli

ng/fix

ture

for

job

8

Inac

cura

te m

easu

rem

ent

of f

ixtu

re t

o m

ount

ing

pla

te

Opera

tor

qual

ific

atio

n6

Too

l "c

rash

es"

as it

com

es d

own

on

mat

eria

l; V

isua

l

insp

ecti

on o

f se

t-up

par

ts

84

83

84

Auto

mat

ed g

aug

ing

that

give

s a

gree

n ligh

t si

gnal

whe

n fi

xtu

re is

alig

ned

pro

per

ly; up

dat

e w

ork

inst

ruct

ions

and

oper

ator

trai

ning

.

Gau

ging

-T.U

.

Aahl

, Y

.R.

Fre

und

z

Doc

.'s

&

Tra

inin

g-E

.V.

Rid

ai

Auto

'd g

aug

ing

inst

alle

d a

nd

valid

ated

5-1

5-

20

XX

; up

dat

ed

WI'

s re

leas

ed 5

-17

-

XX

; al

l op

erat

ors

trai

ned

and

qual

ifed

by

6-1

-

20

XX

83

52

412

0

8O

ne

or m

ore

bol

ts t

oo

loos

e

Opera

tor

qual

ific

atio

n5

Fix

ture

rat

tles

upon

too

l im

pac

t;

Vis

ual in

spec

tion

of

set-

up p

arts

84

03

20

Use

a t

orqu

e w

renc

h to

assu

re p

roper

set

ting

;

updat

e w

ork

ins

truc

tion

s

and

oper

ator

tra

inin

g.

Wre

nch-

B.E

.

Kin

de

Doc

.'s

&

Tra

inin

g-E

.V.

Rid

ai

Tor

que

wre

nche

s

pur

chas

ed a

nd

calib

rate

d b

y 4

-30

-

20

XX

; up

dat

ed

WI'

s re

leas

ed 5

-3-

20

XX

; al

l op

erat

ors

trai

ned

and

qual

ifed

by

5-1

5-

20

XX

83

52

412

0

Med

ium

Ris

k

4. P

erf

orm

requir

ed

PM

Dail

y P

M (

coola

nt

fil

l,

lub

ricati

on)

Low

Ris

k5

. T

urn

machin

e

on,

warm

up

Weekly

/M

onth

ly P

M a

s

requir

ed

Hig

h R

isk

6. S

et

par

amet

ers

Spec

ifie

d s

pee

d,

pre

ssur

e, f

eed

rat

e Pr

essu

re s

et t

oo h

igh

Ex

cess

ive

tool

/fix

ture

wea

r, f

lari

ng o

f pro

duc

t

edge

s re

quir

ing

off-

line

grin

din

g (M

FG

S=

6);

won

't m

ate

to

sub

asse

mb

ly a

t cu

stom

er

loca

tion

(C

ust.

S=8

)

8

Pres

sure

ind

icat

ed b

y

dia

l, d

iffi

cult

to

inte

rpre

t

Opera

tor

qual

ific

atio

n;

Pres

sure

par

amet

er

spec

ifie

d o

n se

t-up

shee

t

7

Opera

tor

slid

es

hand

ove

r par

t ed

ge

to c

heck

for

fla

ring

on s

et-u

p p

arts

85

64

48

Inst

all d

igit

al p

ress

ure

read

-out

; re

-tra

in

oper

ator

s; W

I ad

equa

te

as is

Pres

sure

Inst

rum

enta

tn-

T.U

. Aahl

, Y

.R.

Fre

und

z

Tra

inin

g-E

.V.

Rid

ai

Pres

sure

rea

d-o

ut

inst

alle

d a

nd

valid

ated

5-1

5-

20

XX

; al

l op

erat

ors

trai

ned

and

qual

ifed

by

5-1

8-

20

XX

82

516

80

Specif

ied

mate

rial

per

Bil

l of

Mate

rial

Make p

er

Ass

em

bly

Dra

win

g

fro

m J

ob

Packet

1152

320

3.

Inst

all

tool

ing

&

fix

ture

s

20

XX

-57

2-0

02

Action R

esultsB

.E. K

ind

e

4-8

-20

XX

6-1

2-2

0X

X

Dam

age t

o to

ol/f

ixtu

re,

pro

duc

t ou

t of

spec

. &/o

r

dam

aged

, un

able

to

pro

cess

at

nex

t st

atio

n

Curr

ent P

rocess C

ontr

ols

T

ota

l R

PN

(after

action):

Responsib

ility

&

Targ

et

Com

ple

tion

Date

Recom

mended A

ction(s

)R

PN

CN

R

equirem

ents

Pote

ntial

Failu

re M

ode

Tota

l R

PN

(befo

re a

ction):

Pote

ntial

Cause(s

) of F

ailu

re

Severity

Pote

ntial E

ffect(

s)

of F

ailu

re

Med

ium

Ris

k

Hig

h R

isk

Mis

alig

ned

fix

ture

Cor

rect

ly p

lace

and

align

tool

ing/

fix

ture

s

Pro

cess S

tep

7. M

ake s

et-

up

part

s

Function

System

Process

Activity

Sub-System


Tra

nsactional P

rocess F

ME

A —

Exam

ple

for

Revie

w o

f C

usto

mer

RF

Q

Item

/Opera

tion A

rea:

Cus

tom

er O

rder

Rev

iew

PF

ME

A N

um

ber:

Leve

l:

Page

1of

1

Description:

P

repare

d b

y:

PF

ME

A D

ate

(O

rigin

al):

PF

ME

A D

ate

(R

evi

sed.)

:

Scope:

Core

Team

:

Pre

vention

Occurrence

Dete

ction

Detection

Actions T

aken &

Com

ple

tion D

ate

Severity

Occurrence

Detection

CN

RP

N

Low

Ris

k1.

R

FQ

/PO

receiv

ed

&

confir

med

Com

munic

ate

wit

h c

ust

om

er

Med

ium

Ris

k

2. F

ile c

ust

om

er

data

package

Corr

ectl

y v

ault

data

fil

es

in

cust

om

er

fold

er

on n

etw

ork

inco

rrec

t lo

cati

on u

sed

ne

xt

step

(s)

del

ayed

due

to s

earc

h; f

iles

los

t3

Med

ium

Ris

k

3. R

evi

ew

cust

om

er

data

fil

es

Ensu

re a

ll n

eed

ed

info i

s

receiv

ed

fail t

o ID

mis

sing

/ne

eded

inf

o

nex

t st

ep(s

) d

elay

ed

whi

le inf

o b

eing

pro

vid

ed4

Dete

rmin

e r

equir

em

ents

not

state

d b

ut

consi

dere

d

necess

ary

by u

s.

fail t

o ID

nec

essa

ry

requ

irem

ents

Late

r d

isco

very

cau

ses

del

ays,

ex

tra

cost

,

pro

duc

t fa

ilur

e

8

lack

of

kno

wle

dge

,

over

sigh

t, f

ailu

re t

o

cons

ult

mor

e

kno

wle

dga

ble

per

son,

in

a

hurr

y

Proc

edur

e list

s

thin

gs t

o co

nsid

er

and

opti

on t

o

cons

ult

othe

rs.

6

Sal

es M

gr. r

evie

ws

Quo

tes

bef

ore

they

're

sent

, if

tim

e al

low

s. W

ill

revi

ew m

ore

ofte

n

for

new

peo

ple

.

74

83

36

Dev

elop

S/W

appl

icat

ion

that

guid

es t

he r

evie

w

and

giv

es e

xam

ple

s,

requ

ires

sig

n-o

ff b

y

assi

gned

"fu

ncti

onal

exper

t,"

will no

t al

low

Quo

te c

reat

ion

unle

ss a

ll

fiel

ds

com

ple

ted

.

82

216

32

ID

appli

cab

le

statu

tory

/re

gula

tory

requir

em

ents

.

fail t

o ID

nec

essa

ry

stat

utor

y/re

gula

tory

requ

irem

ents

com

plia

nce

failur

e w

hich

coul

d c

ause

del

ays,

ex

tra

cost

, re

wor

k, la

ck o

f

pro

duc

t ac

cepta

nce,

etc

.

8

lack

of

kno

wle

dge

,

over

sigh

t, f

ailu

re t

o

cons

ult

mor

e

kno

wle

dga

ble

per

son,

in

a

hurr

y

Proc

edur

e

reco

mm

end

s

cons

ulti

ng Q

uality

Mgr

.

4

Sam

e as

ab

ove,

som

etim

es

cust

omer

s cc

Qua

lity

Mgr

.

83

22

56

Sam

e as

ab

ove

82

216

32

ID

pote

nti

al

risk

s to

achie

vem

ent

of r

equir

em

ents

,

request

ed

quanti

ty &

deli

very

date

, etc

.

risk

s no

t id

enti

fied

del

ays,

ex

tra

cost

,

pro

duc

t fa

ilur

e8

lack

of

kno

wle

dge

,

over

sigh

t, f

ailu

re t

o

cons

ult

mor

e

kno

wle

dga

ble

per

son,

in

a

hurr

y

Proc

edur

e list

s

thin

gs t

o co

nsid

er

and

opti

on t

o

cons

ult

othe

rs.

77

56

39

2S

ame

as a

bov

e8

33

24

72

Med

ium

Ris

k5

. C

reate

Quote

Inclu

de a

ll r

equir

em

ent

info

and

any d

evi

ati

on r

equest

s to

requir

em

ents

, quanti

ty,

deli

very

date

, etc

.

Quo

te d

oesn

't r

efle

ct

full r

eality

of

requ

irem

ents

/nee

ds

del

ays

and

ex

tra

cost

7

(ass

umpti

on h

ere

is a

ll

req.

/dev

. inf

o av

aila

ble

)

over

sigh

t, in

a hu

rry

Proc

edur

e ha

s an

exam

ple

of

a Q

uote

form

at in

an

Appen

dix

.

37

21

147

S/W

appl

icat

ion

des

crib

ed a

bov

e w

ill

pop

ulat

e a

Quo

te

tem

pla

te b

ased

on

revi

ew

resu

lts.

72

214

28

Low

Ris

k6

. Sen

d Q

uote

to C

usto

mer

1131

164

Severity

Pote

ntial

Cause(s

) of F

ailu

re

4. R

evi

ew

cust

om

er

requir

em

ents

Outs

ourc

ed

pro

vid

er in

cons

ulta

tion

wit

h in

tern

al

cros

s-

func

tion

al

team

tha

t

incl

udes

Prod

uct

and

Mfg

. Eng

r.,

Prod

ucti

on

Mgr

., Q

uality

Mgr

.,

Purc

hasi

ng

Mgr

.

9-1

5-X

X.

S/W

App v

alid

ated

,

pro

ced

ure

updat

ed,

per

sonn

el t

rain

ed.

Imple

men

ted

10/1

/XX

.

Proc

ess

Mon

itor

ed

and

RPN

reca

lcul

ated

12/1

/XX

.

Sal

es M

gr. r

evie

ws

Quo

tes

bef

ore

they

're

sent

, if

tim

e al

low

s. W

ill

revi

ew m

ore

ofte

n

for

new

peo

ple

.

RP

NR

ecom

mended A

ction(s

)

Responsib

ility

&

Targ

et

Com

ple

tion

Date

Action R

esults

20

XX

-57

2-0

02

W.E

. Kan

dui

t

3-8

-20

XX

6-1

2-2

0X

X

Hig

h R

isk

Curr

ent P

rocess C

ontr

ols

Pro

cess S

tep

R

equirem

ents

Pote

ntial

Failu

re M

ode

Pote

ntial E

ffect(

s)

of F

ailu

re

Tota

l R

PN

(befo

re a

ction):

T

ota

l R

PN

(after

action):

Proc

ess

for

the

revi

ew o

f cu

stom

er R

eque

st f

or Q

uote

(R

FQ

)

Incl

udes

rev

iew

of

cust

omer

req

uire

men

ts f

or n

ew R

FQ

s

Doe

s no

t in

clud

e ex

isti

ng p

rod

uct

or L

egal/

Fin

ance

rev

iew

of

cust

omer

con

trac

t T

erm

s an

d C

ond

itio

ns.

W.E

. Kan

dui

t (S

ales

Mgr

.), O

na

Tim

e (

Insi

de

Sal

es),

Noa

h Pr

ably

m (

Insi

de

Sal

es),

Rei

na M

. Ahlin,

(Q

uality

Mgr

.)

CN

Function

System

Process

Activity

Sub-System

Process FMEA

Documents

PFMEA Workbook - ETI GROUP